Evolution, 4th Edition

MACROEvOLuTiON: EvOLuTiON AbOvE THE SPECiES LEvEL 523

Differences among species in the activity of regulatory genes, such as Hox genes,

or in alternative splicing of key regulatory genes, are sometimes associated with

substantial morphological differences [38]. It is likely that such genes accumulated

their divergent effects gradually, perhaps by successively recruiting different down-

stream genes and pathways. We have seen, for example, that a new spot pattern in

the wing of Drosophila biarmipes resulted from novel activation of the yellow gene by

the Distal-less transcription factor, based on an evolutionary change in the sequence

of a cis-regulatory element (see Fig ure 15.17). Without detailed evidence, we cannot

tell if a big, discontinuous difference in phenotype was caused simply by a single

change in a key regulatory gene, or by multiple changes in the interactions between

the trans-regulator and cis-regulatory changes in many downstream genes.

Some large evolutionary changes do have simple genetic foundations. In some

cases, the gene merely extends or truncates a developmental trajectory without caus-

ing harmful side effects. For example, the few genetic changes that determine the

heterochronic difference between metamorphosing tiger salamanders (Ambystoma

tigrinum) and their paedomorphic relative the axolotl (A. mexicanum; see Figure 15.3)

do not engender an entirely new complex morphology, but merely truncate a complex,

integrated pathway of development that presumably evolved originally by many small

steps. In other cases, as David Stern and Virginie Orgogozo proposed [104, 105], there

has been stepwise, gradual change by the accumulation of successive mutations—but

these mutations have occurred mostly in a single “hotspot” gene that controls a key

point in a developmental pathway. The important feature of such a gene is that it

has few pleiotropic effects on other characters, so mutations are less likely to have

deleterious side effects that would prevent them from increasing by natural selec-

tion. For example, larvae of Drosophila sechellia lack the dorsal trichomes (hairlike

extensions of cell cuticle) possessed by its relatives, such as D. melanogaster (FIGURE

20.5A). The absence of trichomes (a derived trait) is caused by mutations in three

Futuyma Kirkpatrick Evolution, 4e

Sinauer Associates

Troutt Visual Services

Evolution4e_20.05.ai Date 12-28-2016

Input/output

gene

(A) Drosophila sechellia Drosophila melanogaster (B)

shavenbaby

Hox

genes

wingless

soxNeuro Dichaete

EGF-R hedgehog Notch

singed

forked

WASp

shavenoid

CG13913

Arp2/3

enabled

diaphanous

Actin

distribution

miniature

CG15335

CG16798

Apico-basal

complexes

Membrane

matrix

Epidermal cell shape remodeling

Cuticle

yellow

CG17905

Cuticle proteins

Catecholamine

pathway

FIGURE 20.5 How a morphological difference between

species may be caused by repeated evolution in a single gene.

(A) Larval Drosophila sechellia (left) lack the dorsal trichomes

that are present in D. melanogaster (right) and other relatives.

(B) The absence of trichomes is caused by several mutations in

the cis-regulatory control region of the shavenbaby gene, which

determines trichome development in different sectors of each

body segment. This diagram shows the complex gene regula-

tory network, in which shavenbaby is a key player. Transcription

of shavenbaby in different parts of the cuticle is regulated by a

complex of developmental patterning genes (Hox genes, wing-

less, etc.), and shavenbaby in turn regulates the expression of

downstream genes that determine actin distribution, membrane

matrix, and cuticle, which together transform an epidermal cell

into a trichome. (From [105].)

20_EVOL4E_CH20.indd 523 3/22/17 1:44 PM